The Reaction of Oxygen with Carbonaceous Compounds and Trace

18 The Reaction of Oxygen with Carbonaceous Compounds and Trace Mineral Constituents Downloaded by UNIV OF CALIFORNIA SAN DIEGO on July 23, 2016 | http://pubs.acs.org Publication Date: June 1, 1969 | doi: 10.1021/ba-1969-0080.ch018

in an Inductive Electrodeless Discharge CHESTER E. GLEIT North Carolina State University, Raleigh, North Carolina 27607

The relationship between oxidation rate and sample tem perature of graphite in a low pressure oxygen plasma was studied. Energy was supplied from an inductively coupled, 13.56-MHz., 250-watt generator. The apparent activation energy of graphite was 6.5 kcal./mole in the temperature range 100° to 300°C. and zero at higher temperatures. The activation energy of both impure graphite and sucrose were slightly lower. Graphite did not react with a CO plasma below 120°C. and yielded an apparent activation energy of 3.2 kcal./mole between 150° and 300°C. Oxidation rate was dependent on electricalfieldconfiguration. Increasing the temperature of the reaction tube's walls decreased oxidation rate. Trace elements were generally converted to their high est stable oxidation states. Volatility was not strongly tem perature dependent. 2

Τ η 1962 a method was reported for the decomposition of organic sub*· stances based on reaction with an oxygen plasma, which was produced by passing molecular gas through a radiofrequency electrodeless discharge (5). This method has found use in a variety of chemical studies. Loss of trace elements through volatilization and diffusion is less than in conven tional dry ashing. Destruction of mineral structure is reduced. Since a small quantity of purified oxygen is the only reagent, the possibility of chemical contamination is diminished. Applications of this method in clude: microincineration of biological specimens (13), the ashing of coal (6) and filter paper (4), and the recovery of mineral fibers from tissue (J). Several reviews of analytical applications have been pub lished (9,14). 232 Blaustein; Chemical Reactions in Electrical Discharges Advances in Chemistry; American Chemical Society: Washington, DC, 1969.

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on July 23, 2016 | http://pubs.acs.org Publication Date: June 1, 1969 | doi: 10.1021/ba-1969-0080.ch018

18.

GLEiT

Reaction of Oxygen

233

In conventional ashing both reaction rate and retention of volatile components are strongly temperature dependent. However, disagreement exists as to the effect of temperature in plasma ashing. In part, this is attributed to difficulty in applying conventional temperature measuring techniques to solids immersed in a strong radiofrequency field. The question is further complicated by the fact that a single temperature cannot be assigned to the system. Electrons in the low pressure plasma are not in thermal equilibrium with the ions and neutral species. Furthermore, the temperature of the specimen and the surrounding vessel may differ considerably. In this study the effects of temperature on oxidation rate, recombination of active gaseous species, and volatility of inorganic reaction products are considered. Experimental The reaction system employed in these studies (Figure 1) consisted of a 100-cm. long, 3.5-cm. i.d., borosilicate cylinder (Pyrex No. 7740). Specimens were placed on the principal axis 43 cm. from the gas inlet. A 0.6-cm. tube, C., passed diametrically through the cylinder and held the specimen on a central projection. Sample temperature was regulated, in part, by circulating liquids between this tube and a constant temperature bath. With the exception of those measurements in which wall temperature was intentionally altered the external temperature of the reaction vessel was maintained between 20° and 25°C. Forced air cooling was employed. Radio-frequency excitation was supplied from a 250-watt, crystal controlled 13.56-MHz. generator, described previously (3). Except as otherwise noted, an output of 115 watts was employed. Power was transferred to the gas by means of an impedance matching network terminating in a 10-turn coil of 1/4-inch o.d. copper tubing tapped two turns from ground. The coil, B, which had an inside diameter of 4.5 cm. and was 12.5 cm. long, was coaxial with the reaction tube and placed 10 cm. from the gas inlet. After passage through a rotometer, molecular gas entered the reaction system through a capillary orifice, A. Pressure was monitored by a McCloud gauge attached to side arm E, located 50 cm. beyond the gas inlet. Pressure was maintained at 1.2 torr; and flow rates at 150 cc. per minute, S.T.P. U.S.P. grade oxygen and instrument grade carbon dioxide were used. The temperature of solid specimens in the plasma was determined by means of an infrared radiation thermometer Infrascope Model 3-1C00, Huggins Laboratories Inc., Sunnyvale, Calif.) attached to a chart recorder. This device employs a lead sulfide detector and suitable filters to permit remote measurement of 1.2 to 2.5/A radiation emitted by the specimen. Line filters and electrostatic shielding were employed to minimize interaction of the radiofrequency field with the infrared thermometer. To compensate for variations in emissivity and inhomogeneity in the optical field, empirical calibration curves were constructed based on measurements of new and partially oxidized graphite pellets in a con-

Blaustein; Chemical Reactions in Electrical Discharges Advances in Chemistry; American Chemical Society: Washington, DC, 1969.

CHEMICAL REACTIONS IN ELECTRICAL DISCHARGES

234


ventional oven. Thermocouples were employed to determine cylinderwall temperatures. To eliminate interaction with the radiofrequency field, the transmitter was inactivated during the latter measurements. Sequential measurements of both sample and wall temperatures indicated that an extrapolation to generator-on conditions was not required, as cooling rate was low.

Figure 1.

Experimental Apparatus

Capillary inlet; Β = Power coil; C = Specimen mounting tube; D = Sidearm to manometer; and Ε = Outlet to vacuum pump

High purity graphite rods (National Carbon Co., Grade AGKSP) were employed as standard specimens. These rods were 0.61 cm. in diameter and had a cavity machined into their base to affix them to the cooling tube. Pellets of sucrose and carbon containing small quantities of cupric acetate were also employed. The latter were produced by heating and then pressing a slurry of the salt solution and 200-mesh graphite powder. Oxidation rate was determined by weighing the residual solid after 15 minutes. Measurements were usually continued for one hour. Results and Discussion Sample Temperature. The temperature variation of oxidation rate of graphite exposed to the oxygen plasma is shown in Figure 2. Over the range 120° to 300°C., the Arrhenius equation, Κ = C e - a / , fits the data well and yields a value of 6.5 kcal./mole for the apparent activa tion energy, £ . Between 300° and 450°C., the highest temperature studied, oxidation rate is not dependent on temperature. The oxidation of graphite by molecular oxygen at high temperature is strongly de pendent on the purity of the graphite. To determine if a similar effect occurs in plasma oxidation, pellets composed of pure graphite powder E

a


RT

18.

Reaction of Oxygen

GLEiT

235

and inorganic salts were utilized. The results of the addition of 0.01M cupric acetate are shown in Figure 2. At low temperature the change in oxidation rate of pure and impure graphite with temperature is similar. However, the rate of oxidation of the impure graphite becomes inde pendent of temperature at a lower temperature. Measurements performed with specimens containing lower concentrations of cupric acetate led to results which fell between the illustrated curves.


150

r

1

1

1

1

ι

r

Figure 2. Rate of oxidation vs. surface temperature for graphite rods and graphite pellets containing 0.01M cupric acetate Ο—Graphite rods Δ—Graphite pellets

The apparent activation energies are based on the experimentally measured sample surface temperatures. Although this empirically pro vides a satisfactory means of describing the variation in oxidation rate, its theoretical significance must be qualified. The gas in the low pressure electrodeless discharge is chemically similar to that in the positive column of a low pressure arc. Molecule-molecule and ion-molecule collisions are frequent. The ions and neutral species are, therefore, nearly in transla tional equilibrium, with the ions having the greater kinetic energy. Elastic collisions between these species and electrons are less frequent. At low pressure electron temperatures are quite high. In addition, spectroscopic measurements indicate that rotational and vibrational modes are not


236


equilibrated with translational energy. Therefore, no single experimental temperature can describe correctly the reaction system. Furthermore, the number of active gas species is unknown. In the oxygen discharge atomic oxygen ( P) is believed to be the most abundant active species. Higher energy states of atomic oxygen, positive and negative ions, and electronically excited states of molecular oxygen are also present. It is instructive to compare these results with the studies of graphite with atomic oxygen removed from the luminous discharge (2). An early study indicated a value of zero for the activation energy of this reaction (16). Hennig (8) observed no dependence on temperature at a pressure of 1.0 torr and a value of 7 kcal./mole at 10 torr. This difference was attributed to ozone formation. In a recent detailed study of the atomic oxygen-graphite reaction, Marsh et al. (11) reported an activation energy of 10 kcal./mole between 14° and 200°C., which approached zero at 350°C. Although the rate of oxidation in these studies was considerably lower than that obtained in the plasma, their values are consistent with the belief that atomic oxygen is a major reactant within the discharge. Of a variety of specimens, only graphite exhibited an abrupt change in activation energy near 300°C. For example, the apparent activation energy in the decomposition of sucrose is approximately 4 kcal./mole over the entire measurement range. The independence of oxidation rate for graphite at high temperatures is ascribed to the formation of surface oxides. Numerous studies of graphite combustion have shown that such surface compounds control rate at elevated temperatures in excess oxygen. As anticipated, increasing radiofrequency power to the discharge does not measurably increase the rate of graphite oxidation at high sample temperature. The exhaust gas of the oxygen-graphite reaction contains both carbon monoxide and carbon dioxide, with the latter more abundant. In the luminous discharge region these gases react with graphite. The results of measurements in which carbon dioxide was passed through the radiofrequency field and then exposed to graphite rods activated are shown in Figure 3. Between 150° and 400°C. the Arhennius equation fits well, yielding an apparent activation energy of 3.2 kcal./mole. Below 120°C. less than one mg. per hour of carbon was removed. This small weight loss is attributed to ion and electron bombardment. Wall Conditions. In a number of experiments, increasing power beyond an optimum value led to the anomolous result of reducing ashing rate. Earlier it was noted that placing a dry ice-acetone trap in the exhaust stream extended the length of the luminous discharge (5). To study this effect, wall temperatures were varied while the carbon rod was maintained at 200°C. The results of a series of measurements are shown in Figure 4. Assuming that the decrease in oxidation rate at increased


3


Reaction of Oxygen

GLEiT

1.5

2.0 SURFACE

Figure 3.

TEMPERATURE,

2.5 /T

]

xlO

3

Rate of gasification vs. temperature for graphite rods exposed to carbon dioxide discharge

50 40 -

1

1

1

β/®

Ε 111 30 -

0X11 DAT ION RATI


18.

-

20

©

10 1.5 Figure 4.

1

1

. . .

1

2.0 2.5 3.0 WALL TEMPERATURE, k- χ 10

3.5 3

Effect of wall temperature on the oxidation rate of graphite

wall temperature is caused by recombination of active species at the wall, the overall activation energy at low temperatures for this process is approximately 2 kcal./mole.




238

The major sources leading to loss of active oxygen species are recombination of atomic oxygen and ions at the walls of the vessel. The recombination of discharged oxygen on glass surfaces has been investigated. Linnett and Marsden (10) reported that the recombination coefficient of atomic oxygen on clean borosilicate glass was independent of temperature. However, positive values were observed on contaminated surfaces. In a later series of papers Greaves and Linnett (7) studied a number of oxide surfaces, in all cases the recombination coefficient was temperature dependent. For silica apparent activation energies were 1 to 13 kcal./mole over the temperature range 15° to 300°C. The exceptional nature of borosilicate glass was noted. The second major source of loss of active species results from electron-ion recombination at the chamber walls. The electric field restricts the drift of ions to the chamber walls. Rate of recombination is controlled by ambipolar diffusion, but is limited by the presence of negatively charged oxygen ions in the discharge (15). Although diffusion is not dependent on wall temperature, a distortion of the axially symmetric electrical field will enhance wall recombination, producing an increase in wall temperature and a reduction of oxidation rate. This effect is noted in Table I. The presence of a 6-cm. wide metallic ring on the exterior wall of the reaction vessel reduced oxidation rate by more than 10%. Grounding the ring in common with the output coil further reduced the oxidation rate. This is a local effect, in that oxidation rate beyond the ring was not greatly reduced. Distortion of the field can also be introduced by changing the angle between the end of the output coil and the reaction tube. Variations from axial symmetry led to reduced rates and decreased the length of the luminous discharge. Similar effects have been observed in electrodeless discharges at higher pressure (12). Table I.

Effect of Field Distortion on Carbon Oxidation Rate Oxidation Rate

Borosilicate glass tube Borosilicate glass tube and ungrounded copper ring Borosilicate glass tube and grounded copper ring Borosilicate glass tube Borosilicate glass tube

Angle

mg./hi

0° 0°

79 70

0°

62

1° 3°

71 63

Effect on Mineral Constituents. It has been reported that there is no appreciable loss of a number of metal ions in plasma oxidation (4, 5). Non-volatile species include Na(I), Cs(II), Cu(II), Zn(II), Mn(II),


18.

Reaction of Oxygen

GLEiT

239


Pb(II), Cd(II), Co(II), Ho(III), Er(III), Fe(III), Cr(III), As(III), Sb(III), and Mo (VI). The effect of ashing temperature, the reason for the surprisingly low volatility, and the final oxidation state of the product were not previously explored. As part of the present study, 20 to 100 mg. of compounds containing radioactive tracers were deposited on Whatman cellulose filters. After exposure to ashing for a sufficient period to remove the filter paper, generally 30 minutes, the activity of the ash was measured and the oxidation state of the element determined. Specimen temperature was adjusted during ashing by altering input power. These measurements are summarized in Table II. In general, temperature has little effect on retention. These results and earlier observations indicating complete retention of metals in compounds such as arseneous chloride and metalloporphyrins (5) are believed to be caused by competition between volatilization and oxidation to less volatile compounds. Unlike the other elements studied, the highest valence oxide of osmium, Os0 , is the most volatile. Therefore, this element is volatilized in the plasma oxidation process. 4

Table II.

Recovery of Trace Elements from Ashed Filter Paper

Tracer Compound r,9 59

FE

F E

™Se Se "Se Se llOmAg llOmAg 75

75

Fe(OH) Fe(OH) Na Se0 Na SeO. Na Se0 Na Se0 AgCl AgCl Os0 Na 0s0 2

2

3

3

2

2

4

2

4

2

i^Os a

2

4

Percent in Highest Oxidation State

Highest Percent Oxidation Max. Sample State Temperature Recovered 120 120 110 250 110 250 110 250 120 150

100 100 99 100 100 100 89(99)° 79(100)°

The Reaction of Oxygen with Carbonaceous Compounds and Trace

Recommend Documents